Ben Love [00:00:00]:
Good morning, everybody, and welcome to today’s webinar. My name is Ben Love. We’re doing today jointly with John Hollenberg from five by five. How are you, John?

Jon Hollenberg [00:00:09]:
G’day, mate. How are you, mate?

Ben Love [00:00:12]:
I’m living the dream, thank you. I’m actually, actually having a pretty good week at the moment, so I’m celebrating that.

Jon Hollenberg [00:00:17]:
Awesome.

Ben Love [00:00:18]:
All right, so folks, thanks for coming along. Let’s get cracking on this. As I said, my name is Ben Love, I’m the founder and managing director of grassroots it. I will tell you a little bit more about grassroots it in just a minute. Joined here as we saw John Hollenberg from five by five.

Jon Hollenberg [00:00:36]:
Awesome to be here, Ben, appreciate the opportunity, nice work.

Ben Love [00:00:39]:
Thanks, mate.

Jon Hollenberg [00:00:43]:
Do you want me to give you a little intro into who I am and why maybe the team want to listen to me for the next half.

Ben Love [00:00:50]:
An hour, mate, tell us all about John Hollandberg and tell us all about five by five, because the reason we have you here is because you’re an expert in today’s topic, mate, far more than me. So what’s the background?

Jon Hollenberg [00:01:00]:
Well, look, there’s a lot of experts out there. I’m coming at it through the lens of running a digital marketing agency because we’re talking about AI and I just want to share, I guess, our take on that and how we’re utilizing some of these tools and where we see through the world, going from a marketing perspective, looking at through the marketing lens. So a bit of background and context. I operate a digital marketing agency based here at the Sunny Gold coast. Burly heads, we build great looking, high performing digital solutions that generate more leads and inquiries. So very much performance driven approach work with some amazing brands over the years. Here’s a few little sort of snapshots. We’re an award winning digital agency, as I said, based here in Burley, 18 plus years in business.

Jon Hollenberg [00:01:48]:
We’ve delivered thousands and thousands of websites across so many different industries. That’s what I love about this business, is you learn a lot, you go really deep into so many different industry verticals. We’ve grown the team to about 60 team members. So that’s a mix of project managers, developers, designers, content writers, you name it. Really the value proposition here is understanding our clients businesses and being able to distill what can sometimes be quite technical concepts or whatever and bringing that down, a plain english talk. And really what we’re doing is increasing traffic and leads and inquiries. So getting our clients return on investment. What about you, Ben? Who are you and what do you do.

Ben Love [00:02:40]:
I am the founder and managing director of Grassroots it. We are a technology service provider working across it and cloud, cybersecurity and data and automation. Now, we’ve been in business about the same time as five x five. I think we had a funny experience a little while back where John and I were talking, and we realized actually that our 1300 phone numbers that our two businesses use are only a few digits apart. So I think that’s a fair indicator that we actually started our businesses and both went out and registered a 1300 phone number at about the same moment in time. I thought that was pretty funny. We’re based in Brisbane. We have just a bit over 20 people at the moment and growing fairly quickly.

Ben Love [00:03:30]:
And we do support clients and users well, really all around the world because it support and cloud and cybersecurity, it’s really a global game, these.

Jon Hollenberg [00:03:39]:
Awesome, awesome. So, Ben, we are talking marketing in the age of AI. And when we were having a chat about this and preparing, we said maybe it’s a good start would be to define what is AI because we’ve been using AI for years. How would you answer that, Ben?

Ben Love [00:04:01]:
I think it’s a really good place to start. John Tickley so from grassroots it perspective, we actually do spend a lot of time around AI. We’ve got a lot of clients using it. Obviously more and more questions in more recent times coming at us from AI, but it is such a broad topic. Well, that’s the point of today’s webinar really, is to take that very broad topic of AI and really try and focus down on marketing and how AI is impacting and can be used to further your own marketing efforts. But AI itself, as you said, john, AI is actually not new. AI has been around for, I don’t know, a decade or more. I don’t have the numbers in front of me and the years in front of me, but it’s been around for a long time.

Ben Love [00:04:49]:
If we talk about AI as being a computer system which is programmed to behave like a human, to behave like it has intelligence, that’s essentially what we’re talking about with AI. But historically, the types of AI that we’ve been using are really very behind the scenes type things. They’re things to do with machine learning, big data packet, pattern recognition, things like that. So for everybody here who’s ever been on Amazon and purchased something, and down the bottom, you scroll down a bit further and you’ve got all those recommendations where Amazon thinks that you’ve just bought this widget, you might also like to buy these other three things. Those recommendations are being provided by AI. Right? So artificial intelligence in that sense has been around for a long time. The real inflection point, though, that we’ve just hit in the last year or two was, well, it came, of course, with the release of Chat GPT from the company called OpenAI. And that’s a particular type of artificial intelligence, which is called generative AI or Gen AI.

Ben Love [00:06:01]:
And that’s really where the AI is now actually starting to create words, create images, create videos. It’s starting to generate things that didn’t previously exist based on all of these amazing algorithms that have been programmed into it. So hopefully that gives us a little bit of a nutshell to talk about AI today. And I think specifically, John, today, are we focusing on AI or are we going to talk a bit as well?

Jon Hollenberg [00:06:27]:
Yeah, no, it’s more generative, just in terms of looking at it through the lens of us as a digital marketing agency. Like, what are we doing? How are we utilizing these tools? I was reflecting, my dad was involved in it for years. I grew up around computers. When I was ten years old, he came home with this was like pre Windows DOS based operating system, and then loaded up Lotus one two three. And he’s like, this is going to change everything, right? So this spreadsheet, it’s going to save us time in terms of calculating things and doing all the computation and fast forward 30 years. Worksheets or spreadsheets are still very much a thing, and there’s still so much information overwhelm and even more data to process. Right. I see generative AI as like another tool, right? The equivalent of a spreadsheet or something like that.

Jon Hollenberg [00:07:31]:
And just extending on, from what you’ve said, generative AI is here. It’s going to be embedded in the core of our organizations. This is going to be the next productivity revolution. So the question is, how can we set ourselves up to seize on these opportunities? So how’s this going to affect marketing? And it’s very unlikely that anyone on this call hasn’t used some form of AI tools in the last 24 hours or the last week or two. Right? There’s so many tools out there, and it’s literally, as someone thinks of a problem, there’s an AI tool there to solve that problem, whether it’s around content, whether it’s around image generation, whether it’s around videos. And you see the stuff that OpenAI are doing now around video generation and the photorealism. They launched a bunch of videos a couple of weeks ago and it’s insane in terms of just prompt to video output is just wild.

Ben Love [00:08:39]:
It’s remarkable how fast it’s evolving at the moment. I think the speed of adoption of these AI tools of Chat GPT as of course the first one, and then of all of this wave of AI tools that’s coming out. I’ve been in the technology industry for over two decades now. I have never seen anything adopted this quickly. It is truly mind blowing.

Jon Hollenberg [00:09:07]:
Yeah, so then that’s exactly my third point, that just the acceleration of its wild and then how we see this fitting in is speeding up our workflow and execution. This is what we’ve done on a very practical basis here around we create a whole bunch of content for clients. We are leveraging these tools, but it still needs the inputs, the prompts, the quality control, the checks and balances, the human in the mix to make sure that the output is where it needs to be. So what about you, mate? How are you using AI in your day to day marketing? Because not only are you a business owner, you do a lot of marketing, and I observe that how are you leveraging these AI tools in your day to day existence?

Ben Love [00:10:02]:
Probably the main area that we’re using AI in our marketing activities at the moment is really around that content generation piece. There are, I can think of probably two or three main AI tools that we’re using around that and don’t need to get stuck into tools at this stage. But Chat GPT is certainly one of them and a couple of others there, mainly around written content for us. So writing blog posts, writing LinkedIn content and so on. But probably the really key thing to mention here, John, and you absolutely did touch on it there, is that we are not using AI to do all of our writing right. We have not removed the human element from our content creation because following a thought leadership strategy as we are, that still needs to come from the people and from the humans. So we’re really using the AI tools, for example, to help us, I guess, build out the structure of a blog post, for example, what would the overarching structure of this look like? And also, I must admit, I do use it sometimes just when I get stuck, not stuck on the idea or not stuck on the meat of the content that I want to be putting on the page. But simply, how do I write this next sentence? And you get those stupid little writers blocks like that, and AI can really help just fill in all of those gaps that are necessary to build out a full piece of written content.

Ben Love [00:11:45]:
But it’s not actually the meat of the meals. Yeah.

Jon Hollenberg [00:11:49]:
And I’m actually going to run through a practical example of how we do that from a priming perspective. So that’ll sort of be an extension of what you said there. So really, to summarize the impact of AI on marketing strategies, our role as business owners. I’m a business owner, you’re a business owner. We’re also marketers. Right. So we’re tapping into the emotional needs of our customers. Customer has a problem, how do we solve that problem? So tools like chat Chippyt allow us to just use them in a leveraged fashion to work the emotions and with the right prompts, craft that perfect messaging.

Jon Hollenberg [00:12:36]:
But these generative AI tools, they need to learn about what are those problems? And we need to inform it, we need to upskill it. But then that’s then how they can leverage off that based on the vast amount of data points that they have, it’s going to improve decision making. So then you look at just sheer volume of data, and a perfect example here from a marketer’s perspective is things like analytics. Right? So how do I interpret what’s actually happening on my website, on my database, and I don’t have to spend hours and hours setting up worksheets and doing all that stuff. It can very quickly, through the right imports, allow me to make supported decision making in and around trends or consumer behavior or whatever. And really, if we look at different industries, generative AI has been most adopted by marketing and advertising. So there’s a reference in there. Around 77% of marketers are already using this for up to a quarter of their tasks.

Jon Hollenberg [00:13:43]:
Right. And it just essentially allows us to crank out this content. But there’s dangers that come with that, and if it’s not done correctly, you can spot it a mile away. And there’s all these tells. So we’ve got, like, our SEO team have got this rulebook, and every time you see a certain, there’s like five or six certain tell words where you’re just like, that was written by chat GBT.

Ben Love [00:14:15]:
John, I stuck a post up on LinkedIn last week called AI Bengo, and it was asking people to share what the giveaway phrases that they were seeing in written content that says it was written by an AI. And I think my contribution was in the rapidly evolving digital landscape or something to that effect. It’s an absolute giveaway. Like, as soon as you see that, and suddenly so many LinkedIn posts and blog posts are starting with that phrase, and it’s like, oh, come.

Jon Hollenberg [00:14:57]:
Know. And they’ll go into another fact or an argument or something like that. And then reliability. Right, so chat check BTM, there’s a whole bunch of documented instances where it’ll just flat out look you in the eye and lie to you about stuff that is just factually not correct. So it’s sort of like, do your research and make sure what you’re putting up there is technically correct.

Ben Love [00:15:20]:
And that’s the hallucination, isn’t it? That’s how they’ve been referring to that, calling that hallucinating when the ais are giving you things and stating them as facts when they are simply not.

Jon Hollenberg [00:15:31]:
Yeah. And then just the creativity side of things. And this is the danger where if there’s no new ideas in the world, everything just becomes like a mashup. And that’s then where we lack that creativity and just pushing the boundaries and being able to execute on something extraordinary and stifling innovation, essentially. Right? So there’s probably maybe some of the negative impacts. What I might do, I’m going to jump into a practical example, unless you’ve been maybe living under a rock for the last year and a half. I’ll just sort of summarize. Chat JPT, natural language processing tool.

Jon Hollenberg [00:16:16]:
You can have conversations with the chat bot. Like my son will sit there on chat JPT and just have these conversations, literally just back and forward, back and forward, and spend hours doing this. And he’s so engaged with it, which is really interesting. Fastest growing app of all time, trained to follow instructions and prompts, provide a detailed response, free to use. Otherwise you pay your $20 and you get the version four of the engine. Elon Musk says it is scarily good. And then you just see so many tools that are just integrating this. So like even putting together these slide decks, this is using canva, you’ve got these AI assistants where in the little text box it can literally just go and write me some content for a slide or whatever.

Jon Hollenberg [00:17:08]:
Same do with shopify in terms of product descriptions, mailchimp in terms of emails, all that sort of stuff. So it’s all mashed into all these different tools.

Ben Love [00:17:17]:
And John, we’re here at grassroots it. We’re baking Chat GPT and some of these tools into our pre existing tools. So for example, we have got an application here which we use to track all of our client tickets, obviously, and the engineering team works through those tickets and solves client problems. We’re actually have integrated Chat GPT and a couple of the other AI engines into that because it is so easy to do these days using APIs and using the stuff that these platforms make available, it’s there to be used.

Jon Hollenberg [00:17:50]:
And then through Chat GPT you can build your own internal gpts within your business and train that up. And this is sort of like, I guess a smaller example of that. This is just a thread that we use for our content creation. Confirming you can see my screen. Okay, Ben, thumbs up. Awesome. So here’s the priming that goes into creating some content for a specific page on our website for five x five. So we say ignore everything, forget everything, your name is five.

Jon Hollenberg [00:18:22]:
So we’ve got this little alias. You’re an SEO expert that specializes in writing high quality content that converts. We want to create a value and optimize content that is aligned to Google’s helpful content guidelines. We want to make sure that Google likes this content as well. We have over ten years experience in SEO, conversion rate optimization, other forms of digital marketing. Do you understand this? Yes, chat TPT acknowledge that. All right, so hey five, here’s an example of who you write for. So then this gives some background on who we are.

Jon Hollenberg [00:18:57]:
So we’re digital agency award winning. All the stuff that I’ve already told you, we help you share your business magic online, get a stack of leads, all that sort of stuff. Here’s our tone of voice. We are a one stop digital marketing. So we already had some content slot. That content in all of the above, of course, will be reflected in the way that you respond. And then chat JPT says, yes, I’ll make sure that I reflect that and take all that stuff on board. And then I want to test it.

Jon Hollenberg [00:19:26]:
I want to say, hey, can you just reiterate to me, please explain who you are trained, what you are trained on and explain who you’ll be writing for. And then it regurgitates that. So I was like, yes, all right, this is all the stuff that I told you to do. We’re on the right track here. Fantastic. So then that’s the validation that we know that it’s interpreted the information. We say thank you. Now here’s a brief.

Jon Hollenberg [00:19:54]:
We’re going to be writing a page for our pay per click service, so helping clients manage their Google Ads. So please start with an outline of the page. And this is sort of plays to your point, Ben, where it’s like, all right, well here’s the sort of stuff that we want on this page. And you can see automatically because we’ve referenced the Google guidelines, it’s instantly put in a header h one usually the most or second most important element on a page, it’s given us a headline, it’s given us an intro, then it’s given us a h two sub headline. And then the sort of builds out the key points or structure of the page. Give us a call to action as well. Okay, great. Let’s continue with points one, two and three.

Jon Hollenberg [00:20:44]:
One, two and three. And you can sort of see how we chunk it down. So we want to get like a little bit more granular around that rather than just going give me 1000 word article. And then it just goes like that. Right? We want to do it in small incremental sort of chunks. So there’s our headline, there’s our intro, there’s the h two with some content now do four and five. And it continues. We want to expand on some dot points in section four.

Jon Hollenberg [00:21:12]:
So. Okay, cool, that’s good. But we need a bit more meat on the bone there. So then that’s then where you can sort of piece all this together bit by bit and you can sort of see that human prompting and intervention and making sure that every sort of step is steered in the right direction gives us just a much better output and response. And invariably some of this content would have actually been rewritten as well. And you can sort of see how that evolves. We now want some frequently asked questions as well, so we’ve prompted that because we know what sort of frequently asked questions are typically asked and that allows us to do that as well. So this is just like a really practical example.

Jon Hollenberg [00:21:55]:
And then another thing to note is, all right, we’ve already done the hard work in training our bot five on that. Now we’re going to do this for an ecommerce page for SEO services. So you actually don’t have to go back and do that. You’ve just got this big long thread that follows that same sort of process.

Ben Love [00:22:15]:
John, I think this is such an exceptional example of how to use in this instance, Chat GPT. But some of these lessons here are applicable for other AI tools as well. I know a lot of people, myself included, have thrown one sentence in a prompt at something like Chat GPT and then been pretty underwhelmed with the result that’s come back. There’s a lot of talk these days about the importance of prompt engineering, of understanding enough about how Chat GPT works to be able to interact with it, to be able to prompt it in the most effective way, to get the most out of it. And I think your example there, which I absolutely love, it’s one of the best examples I think I’ve seen of prompt stacking or of having this dialogue as you go with the chatbot, with chat. GPT is an exceptional example of really how to get the best out of these tools. Cool.

Jon Hollenberg [00:23:19]:
So that’s an example from a content perspective we look at from a big data perspective. Middle of last year, Google rolled from universal analytics to Google Analytics GA four. Google Analytics four. And for the untrained person, you get in there and you’re like, oh, this feels really clunky. And there’s just a whole bunch of just some learning to do. And it’s very similar in terms of, you just sort of need to be able to navigate your way through that and just change the way that you think about sort of data. So in this instance, I want to see, all right, we’re putting all this social media content out to promote our business. I want to see the impact of that.

Jon Hollenberg [00:24:08]:
So how many people are actually landing on our website? And then how many people are actually engaging in some sort of sales conversation? Right? Are they putting their hand up and say, yes, I’m interested, I need help with this. Right. So that’s then where here, jumping in to Google Geo four, I can literally, it’s like a search engine now where it’s like, how many visits did I get from social media last month? And there’s a whole bunch of different reports and all that sort of stuff. But I can literally just jump in there and see the acquisition sources. Last month, for the month of February, where did the users come from? And I can see organic search, sorry, we have paid social, and these are all obviously the different ways that we’re driving traffic to our site. So I guess the point I’m trying to make here is literally use the search box in GA four to find the information. The other really cool thing that exists here is then where it’s like, you don’t have to be a full data nerd to be able to spot trends. What GA four has been doing increasingly well are these insights and recommendations.

Jon Hollenberg [00:25:27]:
Where I can actually see, on the third of Feb, it says users dropped. We forecasted based on the users that maybe we had historically generated between 47 to 250. And the actual users of 45 was out of that range. Okay, so is there something that happened? Like, did our ads fall over? Was there a server outage? Who knows? Right? So that’s interesting. I can dig into that. And then same deal. There’s a correlation with views dropping. But interestingly, page views of users aged 18 to 24 spiked.

Jon Hollenberg [00:26:11]:
That spiked pretty dramatically.

Ben Love [00:26:13]:
There so.

Jon Hollenberg [00:26:15]:
That’S something else that we can dig into.

Ben Love [00:26:19]:
Maybe that’s your new TikTok strategy starting.

Jon Hollenberg [00:26:22]:
Who knows? Right? But then how can I be alerted for this sort of stuff? So, all right, yes, I will actually want if there’s an anomaly in daily conversions, please let me know about that. I’d definitely like to be to know about that. Maybe daily users as well. This allows me just to pick up on things as and when they happen. So this is an example of, I guess, data being simplified through the use of AI. And then you’ve got all these different insights. It can all be trained and this is just like, you can get lost in here, but want to use it as a practical example to say, all right, don’t know a lot about analytics, but here’s some really simple tools that can just help me keep my finger on the pulse.

Ben Love [00:27:14]:
Yeah, absolutely. I love that. And the whole topic of data analysis and bigger data type stuff is probably a bit outside the realm of today’s discussion. But there are some amazing things that you can do with some of these free tools in that realm. Just as a quick example, I needed to get some insight into some of our help desk stats, so I fed Chat GPT just a CSV dump of some information out of one of our databases and Chat GPT then came back and gave me some analysis and some graphs to really help me get that insight into that data, which would otherwise have taken me probably days trying to slice and dice that in excel or draw some charts or do something. So very powerful tool.

Jon Hollenberg [00:28:03]:
Cool. You’re looking at first ticket response time or average response time. Yeah, cool. Awesome. I like that. Then image creation. So I don’t know who else on the call uses canva, but initially I was reluctant. I started life as a graphic designer.

Jon Hollenberg [00:28:26]:
I spent a lot of hours investing in Photoshop, illustrator, all the Adobe suite. But I will never go back to those tools unless you’re using them for very specific examples.

Ben Love [00:28:38]:
But how long did you fight canva and keep something the table for Indesign and Photoshop and what have you before you caved?

Jon Hollenberg [00:28:48]:
But just like the ability to produce slides and have your templates and have all your assets and actually have them look half decent and all that sort of stuff is amazing. And then you just look at the insane amount of AI tools. They’re probably quite small on the screen here, but you want to cartoonify yourself. You want to create yourself in an emoji, you want to do it in anime. There are just all these infinite tools to be able to do that. So even looking at so this is Dali, we’ll set this up in a let’s use it in existing design, actually, sorry. And look, some of this is a little bit hit and miss. And it’s the same example as Chat GPT, where you’ve got to get really good at prompt engineering.

Jon Hollenberg [00:30:09]:
And this is a whole sub sort of set of prompts around lighting exposure. There’s so many resources out there. Go to YouTube and figure out what the sequence of events looks like. But just even using this as an example, we are generating an image for a horse riding astronaut. Digital art. So the next time that you’re short of an image, maybe you’ve got a blog post, you’ve written, some content, you can literally go and give a good prompt in and around this year and very quickly produce an image that may or may not be appropriate. But that’s probably a little bit more, what is it, impressionist style?

Ben Love [00:31:02]:
It’s the image version of the rapidly evolving digital landscape.

Jon Hollenberg [00:31:07]:
That’s right. And you can see that took me all of like 5 seconds, and I’m on a basic version of canva here. I get 25 credits a month.

Ben Love [00:31:21]:
I think the point here is that that was not an item of clip art that preexisted within canva that John just selected. That is an image which was created from nothing here and now based on that very simple sort of five or six word prompt that John gave it. I think that’s the amazing thing. Know, I tell you what, I wouldn’t want to be running a know library or clipart business or stock photography business or whatever at the moment. I think I would be quite concerned for my future.

Jon Hollenberg [00:31:52]:
While I was chatting to someone about this this morning, about just being a graphic designer, right? And we’ve got five designers on our team. They do a very specialist subset of design, which is user interface design. And that still requires a fair degree of finesse and just creativity in and around a user’s journey and interpreting a brand and all that sort of stuff. But as this evolves, it’s going to be interesting to see what can be automated there. So they’re just like three really practical examples, Ben looking into a crystal ball, future trends and predictions. Who knows what the future like it looks like. But for me, it’s just going to be more and more around this hyper personalization. So the big tech companies, they’re going to know all your deepest, darkest secrets, even if you don’t know them.

Jon Hollenberg [00:32:54]:
Your likes, your interests, your history, and the marketing that’s going to be produced from that is going to be like hyper targeted, that they’ll know that I’m a teenager of the like this obscure californian punk band, and maybe some of that marketing is going to show up with people wearing those t shirts or whatever, right?

Ben Love [00:33:19]:
I’d actually take that a step further, John, and say that I don’t think that’s going to be constrained to big tech companies. I think big tech companies are the ones who’ve had it for a little while and will continue to get even more advanced and possibly creepier with it. But I think we’re actually going to see a lot of this hyper personalization come down into smaller businesses like yours and mine, like your local cafe. I think your local cafe is going to have tech in there which personalizes your experience coming through the door based on everything that they know about you. That’s normally what happens with these things. They start with, it’s like motor car technology starting in Formula One and then coming down to the Toyota Corolla. The same happens with tech. And I think this is moving so quickly that all of this sort of stuff, it’s going to be in the hands of every local business in the next few months like it is moving.

Ben Love [00:34:13]:
So.

Jon Hollenberg [00:34:14]:
Yep, yep. Then we just got the big data side of things, which is the analytics, the interpretation of that data plugging in that helped. This example is perfect in the fact that you’ve got this big data set. It’s like, how can I interpret it? How can I make more informed decisions around my business, how to serve my clients better, all that sort of stuff, and then it’s just content, content, content. Being able to just rapidly just crank the content, that’s just going to be, that’s massive leverage point for every single business here.

Ben Love [00:34:53]:
And John, can I ask you a question about that content piece? Because I really do think this is fascinating in terms of marketing. So when things like chat, GPT, when generative AI have been democratized and we can all access it now for free, which still blows my mind, the sheer volume of content that can be created is phenomenal. So the AI created content and the use of AI in content generation is no longer going to be an advantage. Right? It’s kind of table stakes there. So where will the next advantage come from in terms of content generation? How do you stand out in a content world when all of the content is being created by AI?

Jon Hollenberg [00:35:37]:
Ben, that is an exceptional question because we need to protect the creatives and we need their right brain ideas, because we still need people with ideas that’s really where the magic is going to come from. And the creative person that I still pay a lot of money to here in the organization, I need them because they’re really bloody good at coming up with the ideas we just then got the marketing people, maybe the juniors, whoever, to then go and do the implementation, the execution of that, utilizing these tools through prototyping, through content, through creative work, the repeatable stuff. So I don’t know whether that answers your question or not, but, yeah, I.

Ben Love [00:36:29]:
Think it’s a big question, to be honest. From my perspective, I don’t think we know the answer to that yet, but. But I think it is going to be a very big question.

Jon Hollenberg [00:36:40]:
Yeah. And time will tell, my friend. While we sort of comes to the end of our presentation, we got a few little takeaways and parting shots that we’ll share with everyone. But maybe in the chat, if anyone wants to put any questions or observations or comments or whatever, that’d be amazing.

Ben Love [00:37:04]:
John, we’ve got one from Alina. Do you want to do Q a now or do you want to hit the parting shots first?

Jon Hollenberg [00:37:10]:
Yeah, I’ll quickly do my parting shots and we can do Q A. So really for me it’s just humans are still best equipped to create the A’s ideas and then all we’re doing is leveraging AI to assist in the workflow.

Ben Love [00:37:23]:
About you, mate, this is a great quote that didn’t come from me, but I couldn’t find who it came from. AI is not going to take your job. AI won’t take your job or in this audience won’t take your clients, but somebody who uses AI will. Now, I don’t want that to sound negative and pessimistic. Right. All I’m saying is AI is now table stakes, right? So we need to be adopting AI as table stakes and we need to be looking for the next advantage there. If you don’t adopt AI into your business, I can guarantee that your competitors are and they will have that advantage.

Jon Hollenberg [00:37:59]:
Yeah, I concur. What was the question that you had?

Ben Love [00:38:06]:
Elena. Hello, Alina, how you doing? Thanks for your question. Elena’s written here. Do you think there could be a consumer backlash against the gushing fountains of word salad that generative AI produces or the oily looking fake images? John, totally.

Jon Hollenberg [00:38:24]:
And that’s just then where we need to get better around content. Human intervention, editing, making it fun, providing better prompts, building better detail. Yeah, 100%. And that was just one of the points there where it’s. Yes, there’s just so much junk stuff out there and LinkedIn is just littered with that.

Ben Love [00:38:51]:
From my side. I think the knack, if you will, the trick here that we each have to work out how to address in our own organizations is how to integrate AI into our human processes, right? How to bring the AI tools and the people together for the optimum output. I don’t think it’s a case of just divesting any involvement and delegating that entirely to AI because you will just become part of the word salad, I think it was. Or your words there, Elena. So for me I think it’s actually that integration point. It’s really thinking deeply about how do we use AI without just giving over to. Correct, correct. Thank you, Elena.

Ben Love [00:39:38]:
Great question. I have another question here. This one comes from David. Good morning, David. David writes, how does Microsoft Copilot fit into this discussion? Look, excellent question there. Microsoft were essentially, in some regards co creators of, excuse me. So what they’ve really done is they’ve taken a lot of that AI experience and built it out into what’s actually a suite of products called Copilot. And in very true Microsoft form, they have made it really quite confusing about what Copilot actually means in any particular context.

Ben Love [00:40:22]:
So you can use Microsoft Copilot to do all the same type of things that John and I have just been talking about that you can use Chat GPT for, that’s all there. Copilot is also present in all of the Microsoft Office applications like Word and Excel and PowerPoint and so on. So you can actually work with your copilot AI within those applications to, for example, create a PowerPoint deck. You can use Copilot in PowerPoint and say, hey, copilot, take this annual report Word document I’ve got here and create me a ten page PowerPoint deck that I can use to present to the board. And Copilot will do that for you. The other really deep thing about Copilot though, which none of the other AI tools can do, is that Microsoft Copilot can have access to your corporate data, right? So anything that’s really embedded within your Microsoft three six five tenant or that is in any other systems that you explicitly give it access to, Copilot has access to that data. So it’s not just flying blind, it’s not just using publicly available knowledge, it’s using your corporate data. And that’s really where the unique thing about Copilot is there.

Ben Love [00:41:37]:
But I will get off my little high horse there because grassroots, it is a very heavy Microsoft shop. So the world of copilot pilot is one that I spend a lot of time in. Another question here from Naomi. Hello, Naomi. How do you protect your brand and ideas when using AI, as so many now make the wording so similar. Are there any should not when using your ip within chat, GPT, et cetera? Not entirely sure what that last sentence means, but the crux of the issues there is how do you protect your brand and ideas when using AI, as so many now make the wording so similar?

Jon Hollenberg [00:42:19]:
Yeah, well, I think just reading articles around bigger businesses and you look at team employees and all that, they’re very open to using GPT as a leverage point to increase workflow, productivity, all that sort of stuff. But then it’s like, what sensitive data is being fed into this system? And then how is that being protected? Because there’s obviously future world ramifications around that and all the IP issues and all that sort of stuff. So I don’t know if I’m fully equipped to answer that question, but it’s an, I think there are.

Ben Love [00:43:09]:
If I can have a crack at answering that. Naomi. I think there are two different angles there, or two different ideas, I guess, that your questions really prompted for me, one of them is how do you maintain the purity of your brand in the content and everything that you’re creating without AI watering it down or going a bit off track, or bringing in other influences which really are not core to your brand that you want to be putting out there in the world, whether that be your visual brand or your written brand, voice, or whatever the case may be. Look, I think that comes down to simply vigilance in how you QA and protect your content before it goes out. And that’s a great place for the humans to be, is not simply publishing something that the AI has put out there without a really critical human eye passing over that to protect your brand identity. The second thought that your question prompts for me is really about protection of your own intellectual property, of data protection, data security and so on. Because if you put information into Chat GPT, then there’s a fair chance that Chat GPT will take your information and will ingest that as learning material for itself, and may very well regurgitate it back to some other Chat GPT user at some future point in time. There is a really horrendous case study.

Ben Love [00:44:31]:
I think it was Samsung. They had a software developer do that in the very early days. They used Chat GPT, put in some very confidential something or other from within Samsung, and then, lo and behold, that Chat GPT sort of used that information in a response to somebody not related to Samsung at some later point in time. So, look, that’s really important. And to be honest, that comes down to your understanding of the particular AI tool you are using and what it does with your data. So if you go to the free version of Chat GPT, it will use your data. Right. If you move on to some of the paid versions or use the APIs or you use Microsoft Copilot, they will not use your data.

Ben Love [00:45:12]:
Right. But the point is, you really need to read the fine print on the AI tool that you choose to use so that you know how your data is going to be used when you feed it into this system. We have another question here, excuse me, from Rowan. As our customers apply AI in their own businesses, how do we use robots to market to the robots? John.

Jon Hollenberg [00:45:38]:
Actually, I’m not sure what that means. Well, you think about the process of SEO. We’re using GPT to create articles to impress a robot and get that robot to list that specific page. Already doing it in a directory. Right. But the overarching theme here is make sure you write, and this is the fundamental tenet of Google content guidelines, is like, write it for a human being. Okay. I don’t know if that answers the question or.

Jon Hollenberg [00:46:14]:
I’m not sure.

Ben Love [00:46:15]:
I’m not entirely sure how tongue in cheek’s question might have been.

Jon Hollenberg [00:46:18]:
Yeah, there you go. I think he’s just making a point, but it’s already happening.

Ben Love [00:46:22]:
Look, we’ve got one last question come in from David. This might be our last one. How should a business manager control AI within their own business? AI policy? Risk Qa. Great question.

Jon Hollenberg [00:46:36]:
Yeah. On my to do list, need to navigate that, add it to the list. 100%. Yeah, I think we’re a small marketing team at the moment. Literally like two people. Right. But if you had a big team, you definitely want those policies. You’d want checks and balances, quality assurance, all the above.

Ben Love [00:47:01]:
Yeah. Look, from my side, I think in small business you may be in a position to be a little bit more, shall we say, agile. That’s entirely up to you and how you feel about risk, but certainly the higher the stakes, it becomes a very realistic issue. The whole governance of AI and how you use AI in your business, that comes down to there may be some simple policies there, such as, you can’t put our intellectual property into the AI tool, into Chat GPT, for those reasons we just discussed. But another really interesting example which comes to mind, and I think this was only last week or the week before that, this hit the news. One of the major american US airlines, and I forget which one it was, I’m sorry, but their AI, chatbot, on their website, gave a refund to a customer. The airline then came back and said, actually, we’re not going to give you that refund because it’s against our refund policy. So essentially, the chatbot, for whatever reason, had gone a little bit rogue and had given away money and had given this customer a refund when her policy maybe it shouldn’t have.

Ben Love [00:48:16]:
Now, this made its way into the courts, and the courts upheld that. The chatbot’s refund needed. Needed to be. To be upheld and needed to be given to the customer. Right. So the company was held liable for anything or for what the chat bot had said to their customers. Right. And I’m no lawyer, but to me that seems to be setting a bit of a precedent that I think we all need to be paying attention to managing expectations.

Jon Hollenberg [00:48:43]:
It’s like, on behalf of the company, someone’s been presented with some information and they’ve got an outcome. The company needs to follow through with that.

Ben Love [00:48:54]:
It’s exactly right. But I suspect that came as a bit of a rude shock to whichever airline it was, and everyone decided to duke it out in court. And lo and behold, I think we’ve got a really interesting precedent has been set. So if you have a chatbot on your website at the moment, just think about what it might have access to and might not have access to and what it might be telling your customers.

Jon Hollenberg [00:49:19]:
Love it, mate. Anything else you want to cover off on today?

Ben Love [00:49:22]:
Look, there’s not. That’s the end of our questions. Thank you, everybody who answered those questions. I do appreciate that. Thank you, everybody who attended today’s webinar, of course, feel free to reach out to either John and myself with any questions or follow up that you may have. We are LinkedIn. We are on all the normal places. John, thank you for being here.

Ben Love [00:49:43]:
I greatly appreciate your immense decades of expertise in this area.

Jon Hollenberg [00:49:49]:
Still learning every day. Every day and just scratching the surface. So, yeah, thank you for having me, mate.

Ben Love [00:49:57]:
Fantastic. Thank you, everybody. Have a lovely day.